"Research to Bring More Secure Software for Autonomous Battlefield Operations"
In collaboration with Princeton University, Purdue University will lead research on the protection of Artificial Intelligence (AI) software in drones and other autonomous systems used on the battlefield to save human lives. The project aims to bolster the security of Machine Learning (ML) algorithms applied by these systems for decision-making and adaptation on the battlefield. Researchers will develop a system called SCRAMBLE (Secure Real-time Decision-Making for Autonomous Battlefield) that will be capable of closing the loopholes that could be used by hackers to compromise ML algorithms on different platforms in an autonomous operation. This article continues to discuss the goal, support, development, features, and evaluation of SCRAMBLE.